One quarter of children with cancer do not survive and they experience substantial suffering in the last month of life. For example, over fifty percent of children suffer from pain and dyspnea, symptoms known to be amenable to effective palliation. Retrospective studies have suggested that parent and physician communication during this period is poor, with parents lagging behind physicians by over three months in their understanding that the child has no realistic chance of being cured. Yet, earlier recognition of a child's terminal prognosis by both the physician and parents is associated with better integration of palliative care. Using a prospective cohort design we will perform a pilot project in 117 children with advanced cancer to ascertain parent and physician perceptions about the child's prognosis and treatment goals over time (using a paper and pencil survey instrument). In addition, we will use Pediatric Quality of Life Evaluation of Symptoms Technology (PediQUEST), a hand-held, computer-based system to routinely evaluate quality of life and symptoms experienced by children with advanced cancer. In the intervention phase of this project we will introduce immediate feedback of PediQUEST results to providers. In addition, electronic mail messages will be automatically generated if a patient's symptoms surpass a predetermined threshold. We will preliminarily assess the impact of this system on the natural history of symptoms and suffering, satisfaction, and the timing and quality of end-of-life discussions in order to generate data for sample size calculations for a future randomized controlled clinical trial. Mentored research activities during the award period will be complimented by a career development plan including coursework, seminars, teaching, committee work and clinical responsibilities which will ensure that the candidate makes a successful transition into an independent investigator and becomes a leader in pediatric oncology and palliative care.